Outsmating Malware: Using AI to Combat WannaCry
Marise Miranda, Américo Talarico, Bruno L Sena, Guilherme Paternezzi, Rhian Bottura, Vinicius Benites Bastos, Domingos Sanches · HAL (Le Centre pour la Communication Scientifique Directe) · 2024
Cybersecurity threats like WannaCry are constantly evolving, leaving traditional defenses struggling to keep up. This research explores a powerful new weapon: Artificial Intelligence (AI). Using Bayesian Networks, a type of AI, we developed a method that detects WannaCry with 67.47% accuracy, outperforming traditional signature-based approaches. But this is just the beginning. Our AI model can learn and adapt, making it effective against even unknown malware, offering a crucial advantage in the ever-changing cybersecurity landscape. Keywords: AI-powered detection, Beyond WannaCry, Adaptable defense, WannaCry countermeasures